Abstract

Low-temperature photoluminescence from MoSe₂/WSe₂ moiré superlattice often consists of a broad interlayer emission background with dense, narrow peaks, making microscopic line-by-line assignment difficult. Here, we use hyperspectral photoluminescence mapping and peak-decomposition-free spectral analyses to determine how this spectral complexity is organized in space. A 20 × 20 map acquired with a 400 nm pitch reveals three dominant spectral families that form contiguous real-space domains. Feature-wise spatial correlation analysis and whole-spectrum similarity yield a characteristic micron-scale length of 1.27–2.05 μm, all exceeding the 0.85 μm optical spot size. At the same time, individual pixels retain a dense, multi-peak structure, implying an unresolved local spectral manifold below optical resolution. Correlations among centroid, dominant energy, asymmetry, width, entropy, sharp fraction, and roughness indicate that the micron-scale energy landscape and local manifold complexity can be statistically separated, while remaining correlated across the map, consistent with a hierarchical organization of the emission spectrum. These results establish hierarchical inhomogeneity as an organizing principle of MoSe₂/WSe₂ moiré superlattice photoluminescence.

Abstract

Low-temperature photoluminescence from MoSe₂/WSe₂ moiré superlattice often consists of a broad interlayer emission background with dense, narrow peaks, making microscopic line-by-line assignment difficult. Here, we use hyperspectral photoluminescence mapping and peak-decomposition-free spectral analyses to determine how this spectral complexity is organized in space. A 20 × 20 map acquired with a 400 nm pitch reveals three dominant spectral families that form contiguous real-space domains. Feature-wise spatial correlation analysis and whole-spectrum similarity yield a characteristic micron-scale length of 1.27–2.05 μm, all exceeding the 0.85 μm optical spot size. At the same time, individual pixels retain a dense, multi-peak structure, implying an unresolved local spectral manifold below optical resolution. Correlations among centroid, dominant energy, asymmetry, width, entropy, sharp fraction, and roughness indicate that the micron-scale energy landscape and local manifold complexity can be statistically separated, while remaining correlated across the map, consistent with a hierarchical organization of the emission spectrum. These results establish hierarchical inhomogeneity as an organizing principle of MoSe₂/WSe₂ moiré superlattice photoluminescence.

Abstract

Low-temperature photoluminescence from MoSe₂/WSe₂ moiré superlattice often consists of a broad interlayer emission background with dense, narrow peaks, making microscopic line-by-line assignment difficult. Here, we use hyperspectral photoluminescence mapping and peak-decomposition-free spectral analyses to determine how this spectral complexity is organized in space. A 20 × 20 map acquired with a 400 nm pitch reveals three dominant spectral families that form contiguous real-space domains. Feature-wise spatial correlation analysis and whole-spectrum similarity yield a characteristic micron-scale length of 1.27–2.05 μm, all exceeding the 0.85 μm optical spot size. At the same time, individual pixels retain a dense, multi-peak structure, implying an unresolved local spectral manifold below optical resolution. Correlations among centroid, dominant energy, asymmetry, width, entropy, sharp fraction, and roughness indicate that the micron-scale energy landscape and local manifold complexity can be statistically separated, while remaining correlated across the map, consistent with a hierarchical organization of the emission spectrum. These results establish hierarchical inhomogeneity as an organizing principle of MoSe₂/WSe₂ moiré superlattice photoluminescence.

Abstract

Low-temperature photoluminescence from MoSe₂/WSe₂ moiré superlattice often consists of a broad interlayer emission background with dense, narrow peaks, making microscopic line-by-line assignment difficult. Here, we use hyperspectral photoluminescence mapping and peak-decomposition-free spectral analyses to determine how this spectral complexity is organized in space. A 20 × 20 map acquired with a 400 nm pitch reveals three dominant spectral families that form contiguous real-space domains. Feature-wise spatial correlation analysis and whole-spectrum similarity yield a characteristic micron-scale length of 1.27–2.05 μm, all exceeding the 0.85 μm optical spot size. At the same time, individual pixels retain a dense, multi-peak structure, implying an unresolved local spectral manifold below optical resolution. Correlations among centroid, dominant energy, asymmetry, width, entropy, sharp fraction, and roughness indicate that the micron-scale energy landscape and local manifold complexity can be statistically separated, while remaining correlated across the map, consistent with a hierarchical organization of the emission spectrum. These results establish hierarchical inhomogeneity as an organizing principle of MoSe₂/WSe₂ moiré superlattice photoluminescence.